by Dennis Hong
From the Editor: Dr. Dennis Hong is an associate professor in mechanical engineering and director of the Robotics and Mechanisms Laboratory (RoMeLa) at Virginia Tech University. He and his staff and students have been working with us to develop the blind-drivable car, and he has become a steadfast friend and colleague in our ongoing fight for independence. Following Mark Riccobono’s stirring convention address about the Blind Driver Challenge™ and its place in our struggle for first-class status (See the August/September 2011 issue of the Braille Monitor for the full text of his report), Dr. Hong reviewed the progress he and his team have made in the past year. Here is a slightly edited version of the report he made on the afternoon of July 6 at the 2011 NFB convention:
Good afternoon. Exactly one year ago, at the NFB national convention in Dallas, Texas, I had the honor to speak to you at the convention. I trust most of you here today were there and listened to what I promised you, exactly one year ago. I spoke to you about the Blind Driver Challenge: why we decided to take on this challenge despite all those who told us that we were crazy, how we might be pulling off this seemingly impossible task, and that this is something I truly believe in. I also promised you that we would complete our first prototype vehicle and demonstrate a blind person driving it safely and independently--not being driven, but driving it by making active decisions. And we have kept our promise. On January 29 this year, on that historic day, Mark Riccobono drove the 1.5-mile course on the famed Daytona International Speedway, dodging obstacles and even passing a vehicle, in front of a huge crowd and media representatives from all over the world. What an exciting day it was!
However, I think the real excitement begins now. I have kept my promise of a year ago to develop a vehicle, and today I promise you that this is just the beginning.
When we first announced that we were going to accept this challenge, many thought we were crazy, most of them doubted that it could be done, and even some of my colleagues challenged us on the whole idea, asking why we had decided to develop a vehicle for the blind. Setting aside the ridiculous claims that this technology and the demonstration were all fake, there are actually some genuine arguments, legitimate criticisms, realistic concerns, and valuable questions worth thinking about. Let me mention a few of them.
First, many voiced their concerns about the safety of this system. Some of these concerns arose because the critics did not trust the technology, some because they do not trust the capability of the blind. I won’t comment on the latter since you know better than I what you can and cannot do. But with the right nonvisual user interfaces, once we can deliver all the information the driver needs to operate the vehicle safely, I believe that the blind can perform equally, possibly better than the sighted. As a matter of fact, we were surprised to find that our initial testing showed that the blind performed better with this vehicle than the sighted! So officially Mark Riccobono is a better driver than I am, at least in this vehicle.
Regarding doubts about the safety of the technology, this concern has some validity. No system is fail proof. However, this is where true engineering comes in. I would like to ask those who worry about the system’s failing--when was the last time you doubted the autopilot that most likely was flying your plane during your last flight? But to reassure those who doubt the safety of the system, I would point out that the vehicle won’t be on the public roads until proven as safe as, or safer than, today’s vehicles for the sighted, and I believe this can be done.
Second, some worry that, even when the technology becomes mature enough and safe enough, the vehicle will be too expensive for people to afford. This is another reasonable concern and an important one. We started this project to give freedom and independence to the blind, but, if only a few people could afford it, what would that accomplish? This is actually a common problem for consumer products that use cutting-edge technology. But we believe that mass production will lower the cost, just as with personal computers, mobile phones, and the standard automobiles of today. Still, this vehicle will be more expensive than a standard vehicle, so maybe a government subsidy program would be needed to make it more affordable for the masses.
Third, Some people doubted the type and amount of information being received by the driver from the nonvisual user interfaces we developed for the vehicle. Some said “We (sighted people) don't trust the safety of the technology. Driving involves so many things beyond what the other senses can handle that this approach is inherently dangerous.” Others said, “With DriveGrip and SpeedStrip it's just a blind person doing what the electric motors and actuators are doing in an autonomous vehicle. This is not freedom for blind people.”
Who is the best person to answer these questions? Just ask Mark Riccobono! (I always delegate the difficult questions to Mark, anyway.)
But seriously, as Mark mentioned in his speech, so much more of the other senses is needed than is required by those nonvisual user interfaces. Mark made active use of his other senses such as listening to the sound the tires generate rolling over the roads, exactly as sighted drivers do. He used the sense of angular and linear acceleration perceived by his inner ear to make judgments and adjustments to his control of the vehicle, exactly as sighted people do.
If you think about it, it is really just that the blind cannot receive information using sight. Other than that, blind drivers are in exactly the same situation as sighted drivers since they can use all of their other senses, and, most important, use their brains to make decisions just like all other drivers, and what’s wrong with that?
As for the type and amount of information transferred to the driver, we are now researching and developing the generation of nonvisual user interfaces beyond DriveGrip and SpeedStrip, which we hope to demonstrate on the vehicle in the near future.
If one can believe that we will have fully autonomous, driverless cars in the future, which most people do, there is absolutely no reason why one should doubt the possibility of a car safely driven by the blind on public roads. If someone does doubt this eventuality, that means that person doubts the capability of the blind and needs a serious kick in the butt.
Last, some argued that, by the time this technology becomes mature, we will have fully autonomous, driverless vehicles driving us around, so we don’t need this technology. Interestingly, I also got some emails from blind people saying that they do not want this technology, that they would rather have better public transportation and wish we would put more effort and money into fully autonomous vehicle technology instead because they don't want to drive and cause an accident and prefer being driven safely.
This is probably the most difficult and controversial question to answer, but it is an argument worth thinking about. If we will have driverless cars in the future, why are we spending time, money, and effort in developing a car that a blind person can drive? Well, I can think of many reasons. Wouldn’t you like to experience the freedom and independence of driving your own car like the sighted? But maybe equally important is the potentially huge value of the spin-off technologies we develop through this project. The nonvisual user interfaces we develop can be used for applications other than for driving--home appliances, office equipment, and educational technology. The possibilities are endless. We would also like to show the world the true capacity of the blind through this groundbreaking project. Let people know the truth about blindness just as I have learned while working with the NFB on this project. I also have a personal reason for pursuing this--as an educator I want to inspire other scientists and engineers to develop new technology to help the blind, and I believe we are succeeding in these areas.
Now for a few of my thoughts about developing technologies for a driverless, autonomous car verses a car for the blind. Surprisingly, the two cars share many of the same technologies. The sensors we use for the blind driver challenge vehicle are almost identical to the ones we use for autonomous vehicles. Moreover, in this project we used many of the technologies we developed for the autonomous cars we used in the 2007 DARPA Urban Challenge. But the similarity ends there. The focus of autonomous-vehicle research is developing intelligent vehicles (AI) for cars, artificial intelligence in some sense, while the focus of the Blind Driver Challenge™ vehicle is developing methods to convey a vast amount of information to the driver fast enough and accurately enough for safe driving using nonvisual means. As a matter of fact, one of the reasons we chose driving as the application is that driving is an activity that uses and relies on vision the most; so, if the interfaces work in driving, we believe that it can be used for almost any other application for the blind as well.
For those who think we should focus our efforts only in developing technologies for driverless cars, remember that, even if you ride in an autonomous vehicle, you will want and sometimes need to switch to the manual mode from time to time. This will be true for both blind and sighted drivers. In some situations the computer won’t be able to handle the demands, and setting aside necessity, the sheer fun of driving requires this interface. I am sure that some day in the not too distant future I will purchase an autonomous car for my family. Even though it will be capable of driving me to my destination, I will want to take over the wheel and drive it from time to time, just for the love of driving. Thus I believe both kinds of research are important, and we can all benefit from both.
As you may already know, this groundbreaking project has generated a huge buzz in the public media. The story has been featured on magazine covers, in newspapers, in hundreds of blogs on the Internet, and on television news programs, both national and international. Reporters and TV crews from Germany, England, Korea, Russia, and China have been following us. It has been dizzying. Just yesterday a TV crew from the largest network in Russia was here to interview Mark and cover the story of the Blind Driver Challenge project. Diane Sawyer from ABC News will be coming to the NFB Youth Slam in 2 weeks to cover the story.
However, the single event that has had the largest impact on this project was probably a talk I delivered in March at a conference called TED. This conference is relatively new. It covers topics on technology, entertainment, and design. It is considered one of the highest-impact talks one can give. Previous speakers have included Bill Gates, Steve Jobs, Al Gore, and many of the technology movers and shakers, including our own Ray Kurzweil. As a matter of fact, I was originally invited to give a talk about my other research work on humanoid robots, but I decided to change the topic to the Blind Driver Challenge just two weeks before the conference because I wanted to seize this opportunity to share this important work with the world. I actually ran into some big trouble because of this decision, but it was worth it. Since my talk was posted online on <TED.com> just a few weeks ago, it has already received more than 250,000 views online. This is important for our mission to spread the idea and to let the world know.
So where do we go from here? What is the future of the Blind Driver Challenge vehicle? What is our next step in developing technology for the blind? What is the next challenge that we will accept?
As for the Blind Driver Challenge vehicle, instead of the car itself, we will be focusing our efforts on developing new types of nonvisual interfaces. We have decided to take this approach for several reasons. First, we believe the vehicle itself is mature enough to be used as a research platform as is for many years to come. We have developed two identical vehicles and want to share the technology with other researchers and invite them to join in collective efforts. Second, we would like to leverage the work of other researchers working on autonomous-vehicle technologies because we can use the results from their research instead of duplicating efforts. Third, by the time the system is mature enough, we believe that the infrastructure for public roads will have changed. For example, even though our current vehicle cannot see the traffic lights or read traffic signs, we believe that intelligent roads are coming soon, so communication between traffic signals and vehicles will become a reality relatively soon. Last, we believe the highest impact will come from the development of such interfaces, which can be used not only for driving, but also for other areas as well.
So what is our next challenge, a difficult challenge that we accept? I talked to many of you at last year’s NFB convention, and many people expressed the need for an indoor navigation device. I have seen a number of good solutions available today for outdoor navigation using GPS technology, that I assume you already know about. However, since GPS relies on satellites, this approach does not work indoors. To gain true independence, one needs the ability to navigate challenging environments such as shopping malls or airports. At a mall how do you find where you are and where to go to find your favorite pair of jeans at the Gap? In a grocery store that you have not visited before, how do you find the aisle that has your favorite brand of cereal? Even at this convention, if you need to go to the restroom and return to your table at the banquet, how do you find your way back? We have already been quietly working on this technology for the past few months, and we have developed a very early proof-of-concept system that has the potential to make this a reality. We still have a long way to go, but I am glad to tell you that our initial results are very promising. Hopefully we will be able to demonstrate this technology within a year or two.
I have kept the promise I made exactly a year ago of developing a car for the blind. As a matter of fact, this was really NOT a demonstration of the cutting-edge technology we have developed, but rather a true demonstration of the capacity of the blind, what you can achieve with a little help from technology.
The Blind Driver Challenge™ was not about the car, not about driving. It was about the interfaces and technologies that can change the world, and, more important than that, it is about shattering false perceptions of blindness.
As I have already mentioned, I promise that this is just the beginning. Expect to see amazing things and technologies coming soon that can change the world. Dr. Maurer’s vision is closer to reality than you might think. Thank you.